Goal announces to Llamasacon, its first generative conference of the development of AI

Goal on Tuesday announced which will organize its first development conference dedicated to generative AI.

CALLACON CALL After the family of the Metrativa Meta’s calls, the conference takes place on April 29. Goal said he plans to share “the latest in [its] Open source AI developments to help developers […] Create incredible applications and products. “

Additional details will be available soon, Meta said. The company’s annual conference of developers, Meta Connect, will take place later in the year, in September, its typical window.

Several years ago he adopted an “open” approach to develop AI technologies in an attempt to promote an applications and platform ecosystem. It has not been revealed how many applications or services have been built on top, but previously it has been observed that Goldman Sachs, Nomura Holdings, AT&T, Dordash and Accenture use flame. The company claims to have hundreds of millions of downloads of the model, and at least 25 members that house call, including Nvidia, Databricks, Groq, Dell and Snowflake, some of whom have created additional tools that, for example, allow the models to be Reference proprietary data and allow them to function in lower latencies.

But, according to the reports, goal was trapped flat due to the emergence of the Chinese company of the Deepseek, which managed to launch the “open” to rival the goal itself.

ReportedlyMeta believes that one of Deepseek’s newest models could overcome the next flame version, which will be launched in the coming weeks. It is said that Meta hastened to establish war rooms to decipher how Deepseek reduced the cost of running and deploying models, so that they could apply those learning to the flame development itself.

Recently, Meta said that I would spend up to $ 80 billion on AI -related projects this year, including AI hiring and the construction of new AI data centers.

Meta Mark Zuckerberg CEO previously announced That the company plans to launch several flame models in the coming months, including “reasoning” models in the OPENAI O3-mini line and models with multimodal capabilities. He has also suggested “agents” capabilities, suggesting that Future flame models will be able to take certain actions autonomously.

“I think this could be the year that calls and Open Source become the most advanced and widely used models,” Zuckerberg said during the earning call of the fourth quarter of the goal 2024 in January. “[O]Your goal for [Llama this year] It is leading. “

Goal is also in the midst of a demand that accuses the company of training its models in book materials with copyright without permission. In another challenge for the ambitions of the Meta flame, several EU countries have forced the company to postpone, and in some cases cancel completely their model launch plans on data privacy concerns.

Techcrunch has a Bulletin focused on AI! Register here to get it on your entrance tray every Wednesday.

#Goal #announces #Llamasacon #generative #conference #development

Leave a Reply

Your email address will not be published. Required fields are marked *